Geolocation Methods: A guide to successfully collecting broadband deployment data

Summary

This Universal Service Administrative Co. guide tells carriers in modernized Connect America programs how to collect latitude and longitude coordinates for the broadband deployment data they must file in USAC's High Cost Universal Broadband (HUBB) portal. It describes three methods: geolocation with GPS in the field (best results), desktop geolocation using web-based maps and imagery (very good results) and automated address geocoding (good results). Coordinates should be taken inside the footprint of the structure served, never at the network node or pedestal, and reported in decimal degrees to at least six decimal places. The geocoding section warns that rural interpolation can place points on the wrong side of the street, and that the HUBB will reject points matching ZIP Code centroids. A table compares the methods by data confidence, effort, pros and cons.

Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used

Full text

Geolocation Methods
A guide to successfully collecting broadband deployment data

Carriers participating in modernized Connect America programs must file complete, reliable
geolocated broadband deployment data with USAC’s High Cost Universal Broadband (HUBB) portal.
This data will show where you are building out mass-market, high-speed Internet service. To help you
prepare, we have created this guide describing three methods for gathering geolocation data (latitude
and longitude coordinates), with the pros and cons of each approach.

You can use any method or combination of methods to gather the correct coordinates for the
locations you serve. Be mindful that accurate data is key to filing successfully with the HUBB. This data
will show the impact of Connect America funding and will appear on a public-facing map to display
progress in expanding broadband throughout rural America.

Contents
• Overview of Geolocation Methods
• Geolocated Data Collection Guidance
• Three Methods:
o Geolocation with GPS in the Field
o Desktop Geolocation Using Web-Based Maps and Imagery
o Automated Address Geocoding
• Comparison of Methods

1
Overview of Geolocation Methods
Geolocation is the process used to collect latitude and longitude (lat/long) coordinates that represent a specific
location. We describe three geolocation methods:

• Geolocation with GPS in the field. (Best results)
This process involves going out in the field, standing at a location with a smartphone, tablet, or
other mobile device enabled with GPS technology and capturing the lat/long coordinates.

• Desktop geolocation using web-based maps and imagery. (Very good results)
This approach uses secondary resources, such as web-based maps and imagery, to identify
specific locations and look up the corresponding lat/long coordinates.

• Automated address geocoding. (Good results)
This method, which can be used to process large batches of street addresses, uses an address
geocoder program to convert addresses into lat/long coordinates.

Geolocation Data Collection Guidance
When collecting geolocated deployment data for the HUBB, follow this guidance:

• Whenever possible, location coordinates should represent the structure being served and should
be collected at some point inside the structure’s footprint, also called a “rooftop” coordinate.
Ideally, you should gather coordinates at a spot unambiguously associated with the structure,
such as the network connection point or the front door.
• At a minimum, coordinates should represent a point on the correct property or parcel of land
where service is being delivered. You should try to collect coordinates as close as possible to the
structure being served.
• Do not submit coordinates at the network node or pedestal used to serve a location.
• For all location data collection methods, latitude and longitude coordinates must be reported in
decimal degree format, to at least six decimal places. For example, the coordinates for the USAC
office in Washington, DC are:

Latitude: 33.898618
Longitude: -77.028845

Universal Service Administrative Co. 2
Geolocation with GPS in the Field
The most precise method for collecting lat/long
data is to go into the field and capture
coordinates at each individual location using a
GPS-enabled device. This method will produce
the most reliable results with the fewest errors,
ensure that coordinates are gathered for the
correct location, and reduce the need for
additional data collection, verification, and/or
clean-up.

Recommended Collection Method
There are several easy-to-use GPS apps on the market that you can install on a smartphone or tablet either for
free or at a very low cost. Equipping your installation, inspection, and repair crews already in the field with GPS-
enabled devices can minimize the need for specific trips to collect data.
When collecting geolocations in the field, make sure you capture the lat/long coordinates from a point on the
correct property or parcel of land where service is actually being delivered. Ideally, you should capture
coordinates at a point within the footprint of the actual structure being served, such as the network connection
point or the front door.

Things to Consider
Field work can be time-consuming and expensive. To be as efficient as possible using this method:

• Outfit field crews with GPS technology to collect location coordinates during routine
maintenance, repair, and construction,
• Minimize technology issues by performing setup, testing, and any needed training before heading
out in the field, and
• Plan the field work in advance. Determine an optimal route, or a set of locations to be collected
during the same trip, to make the most effective use of time and resources.

Universal Service Administrative Co. 3
Desktop Geolocation Using Web-Based Maps and Imagery
Desktop geolocation uses secondary resources, such as web-based maps and imagery, to identify specific
locations and look up the corresponding lat/long coordinates. Although this is typically a manual process, it can
produce accurate results, assuming that the web-based maps and imagery are good quality and that the
location can be properly identified.

Recommended Collection Method
Below is a typical workflow for collecting lat/long coordinates from a web-based map. Please note that there are
many different methods and resources for this type of process.
1. Open a web-based map and navigate to the location you want to geolocate. This can be done
by either searching for an address or zooming and panning in the map.
2. Turn on the imagery (i.e., aerial photo view) and zoom in to the specific location/structure.
3. Once at the location, identify the lat/long coordinates within the application. (With many
applications this requires placing the cursor over the location and right clicking).
4. Copy and paste the lat/long coordinates into the appropriate columns in your file for
uploading to the HUBB.

Things to Consider
When it is not feasible to go into the field to visit network locations with a GPS device, desktop geolocation offers
a viable alternative. This approach can be highly reliable, but be aware that:
• This method relies on available maps and imagery, which in some cases may be out of date,
particularly in rural areas. Finding locations of new structures, for instance, can pose a challenge
if the only available maps and images predate construction. These types of situations introduce
uncertainty and may require verification using another geolocation method.
• It may be difficult to identify specific structures if there is a lot of tree cover or if other nearby
buildings have the same footprint and look very similar from above.

Universal Service Administrative Co. 4
Automated Address Geocoding
Geocoding is an automated method that uses an address
geocoder program to convert address information into
lat/long coordinates. This approach can process large
batches of addresses in bulk and at a low cost. But the results
can vary in accuracy, based on the quality of the address
input data and the underlying reference data.

Geocoding programs work by deciphering the parts of an
address and attempting to provide matching lat/long
coordinates based on underlying reference data. The
reference data may contain an exact match to an address,
producing a dependable result. But in many cases,
particularly in rural areas, geocoding programs may not be
able to find a match for every address and instead
“interpolate” (or estimate) coordinates, leading to less reliable
outcomes. In some cases, geocoders may not be able to find
matches at all if locations are missing from the underlying
reference data.

At times, geocoding can even produce coordinates that are
far removed from the actual location being mapped,
particularly in rural areas where the reference data is often
unreliable or simply out of date. This means that you will
need to review and clean up geocoding results – and
potentially reprocess some of them – to ensure accuracy.

Recommended Collection Method
There are several different types of geocoders on the market that can process large batches of data, including a
number of commercial products. Costs and data output quality vary.

A number of state GIS (geographic information systems) agencies have also published free geocoders.
Government agencies are continually working to improve address data for 911 emergency response purposes
and therefore often have high-quality geocoders. Here is a partial list of available, government geocoders:

State Geocoder URL

Arkansas http://gis.arkansas.gov/?product=arkansas-statewide-geocoder

Delaware https://firstmap.gis.delaware.gov/pdfs/public/FirstMapGeocoder.pdf

Universal Service Administrative Co. 5
Indiana https://gis.in.gov/arcgis/rest/services/Indiana_Composite_Locator/GeocodeServer

Montana http://ftp.geoinfo.msl.mt.gov/Documents/Applications/UsingMSLGeocodingService.pdf

New Jersey https://njgin.state.nj.us/NJ_NJGINExplorer/jviewer.jsp?pg=Address

New York http://gis.ny.gov/gisdata/inventories/details.cfm?DSID=1278

Utah https://gis.utah.gov/using-the-mapserv-utah-gov-api-to-geocode-address/

Vermont http://vcgi.vermont.gov/warehouse/web_services/geocoding

Washington, DC http://octo.dc.gov/service/master-address-repository

Things to Consider
The output of geocoders is only as good as the address input data and the underlying reference information
(which varies by geocoder). Here are some things to keep in mind:

• Be sure that addresses are accurate, complete, and consistent before submitting them to a
geocoding program. Missing or incorrect ZIP Codes, for example, can lead to serious errors.
• Be aware of geocoder limitations. Automated processes often have difficulty handling certain
situations, such as:
o Ambiguous addresses
(i.e., 214 Elm Terrace vs. 214 W Elm Terrace)
o Inconsistent address formatting
(i.e., 742 West Main Street vs. 742 W Main St vs. 742 Main St W.)
• Keep in mind that rural addresses can be difficult to accurately geocode since reference data may
be unreliable or incomplete. When this is the case, geocoders typically use an interpolation
process which assigns lat/long coordinates to a location along a street segment based on street
address range. For instance, if the addresses along a length of street range from 100 to 200 and
the address your geocoding is 150, the geocoder will place the location halfway down the length
of the street. This approach can lead to incorrect results, such as:

o Lat/long located far from the actual location. The geocoder’s reference data may include the
maximum range of potential addresses on a street segment (1-99) even if there are fewer
actual addresses assigned in the real world (57-89). This can result in the interpolation of an
address to an incorrect point along the road and possibly in an area not eligible for support.
o Lat/long located on the wrong side of the street. Geocoders often assume that all even-
numbered addresses are on one side of the street and all odd-numbered addresses are on
the other side of the street. But this is not always the case in the real world, so the resulting
coordinates may be on the wrong side of the road and possibly in an area not eligible for
support.

Universal Service Administrative Co. 6
o Cascading to less granular geographic levels. If the geocoder is unable to make a match to a
street address, it may cascade to a match at the ZIP Code level and assign lat/long
coordinates for the centroid of the ZIP Code area. The HUBB will reject points matching ZIP
Code centroids.

Comparison of Methods
The diagram below shows the general relationship
between the reliability of data output and the level
of effort put in for each of the three geolocation
methods. In short, the best data requires more up-
front effort but pays off in the long term with
reliable, high quality data.

This table summarizes the differences between the methods.

Method Data Confidence Level of Effort Needed Pros Cons

Geolocation with Best Results: Highest: Best accuracy and Time consuming to travel
GPS in the Field Yields extremely reliable Coordinates are captured highest confidence that to each location and
and highly consistent by physically traveling to the lat/long represent requires some minimal
results since the data the location, although the actual deployment setup/training for data
collection is performed at effort is reduced if location collection via mobile app
the location in the field coordinates are collected
when crews are already in
the field

Desktop Very Good Results: Output Medium: High confidence in Manual process to find
Geolocation using is generally very reliable, Coordinates are captured results and you only each location on map,
Web-based but depends on quality of by reviewing imagery on a need a computer with and then click and
maps/imagery imagery and ability to desktop/laptop and an Internet connection capture coordinates for
properly identify locations placing a point for each each one
on a web-based map location

Automated Good Results: Lowest: Able to quickly process Results may vary in
Address Geocoding Results depend greatly on Addresses are large batches of data completeness and
the quality of the address automatically geocoded accuracy, and some may
input data and the in batches, although not match actual
underlying reference data results may need to be location. May need
reviewed and potentially software and expertise to
re-run or field verified run the geocoding
(which would require process. Data clean-up
more effort) and verification may be
needed after addresses
are geocoded

Universal Service Administrative Co. 7
Examples
The diagrams below display two different rural locations, and illustrate how the methods described above can
produce different lat/long coordinates.

Example 1: An address in Vermontville, NY 12989

Example 2: An address in Saranac Lake, NY 12983

Universal Service Administrative Co. 8

File and source

File: HUBBGeolocationMethods.pdf
Size: 394,876 bytes
SHA-256: a59ff33ae77710ca6cd11be3064a8c88115451b75ae467a68c69c9890a26179d
Our copy: HUBBGeolocationMethods.pdf
Original: www.usac.org